Abstract 3779: Genomic and clinical characterization of metastatic patterns using real-word data from a large cohort of colorectal cancer patients

Abstract

Abstract Background: Colorectal cancers (CRCs) can exhibit disparate metastatic patterns that affect outcomes. We investigated associations between genomic features and organ-specific metastatic tropism and we compared the frequency of actionable alterations detected in patients undergoing longitudinal clinical sequencing. Methods: We analyzed data for 6574 tumors from 6153 CRC patients, including 4510 primary and 2064 metastatic specimens. All patients had microsatellite stable (MSS) disease and were treated at Memorial Sloan Kettering. Paired primary and metastatic samples were available for 203 patients. Tumors were sequenced with MSK-IMPACT, a targeted DNA sequencing assay that identifies genomic alterations in 341-505 genes. Clinical annotations included sex, stage at diagnosis (Dx), and primary tumor location. History of metastatic disease was automatically retrieved from radiology reports using validated natural language processing methods; time to specific-site metastasis was modeled with Cox regression. False discovery rate (FDR) was used to correct for multiple hypothesis testing. The OncoKB knowledge base was used to identify clinically actionable alterations. Results: Liver (n=1068, 51.7%), peritoneal (n=269, 13.0%) and lung metastases (n=263, 12.7%) accounted for the majority of non-primary sequenced specimens. The frequency of actionable alterations (OncoKB Level≥3A) did not vary between unmatched metastatic and primary samples (12.3% vs 11.4%, p=0.302). In primary tumor samples from patients with localized disease at the time of diagnosis, TGFB pathway alterations were associated with increased risk of metastasis to the brain (HR=2.41, p=0.028), bone (HR=2.10, p<0.001), peritoneum (HR=1.82, p=0.001) and liver (HR=1.42, p=0.049). Alterations in the RAS pathway were associated with increased risk of lung (HR=1.47, p=0.009), peritoneum (HR=1.54, p=0.021) and liver (HR=1.44, p=0.026) metastases. By contrast, WNT pathway alterations were associated with a lower risk of developing lung (HR=0.69, p=0.033) and peritoneal metastasis (HR=0.68, p=0.081). Among patients with paired primary and metastatic samples, no significant difference was observed between the frequency of clinically actionable alterations in primary vs. metastatic specimens (16.7% vs. 18.2%, p=0.794). Presence of metastases-private or primary-private alterations did not correlate with site-specific metastasis risk. Conclusions: The risk of developing metastatic disease and organ-specific dissemination might be linked to genomic abnormalities detectable at the localized stage. No significant difference in the frequency of actionable alterations was observed between primary and metastatic CRC samples. Citation Format: Paolo Manca, Ayush V. Kris, Henry Walch, Christopher Fong, Justin Jee, Karl Pichotta, Nikolaus Schultz, Walid K. Chatila, Rona Yaeger, Francisco Sanchez-Vega. Genomic and clinical characterization of metastatic patterns using real-word data from a large cohort of colorectal cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3779.